(19) TZZ__T

(11) EP 1 964 557 B1

(12) EUROPEAN PATENT SPECIFICATION

(45) Date of publication and mention (51) Int Cl.: of the grant of the patent: A61K 31/285 (2006.01) A61K 33/36 (2006.01) 02.01.2013 Bulletin 2013/01 A61P 35/02 (2006.01)

(21) Application number: 08075200.9

(22) Date of filing: 10.11.1998

(54) Process for producing trioxide formulations Verfahren zur Herstellung von Arsentrioxidformulierungen Procédé de production de formulations à base de trioxide d’arsenic

(84) Designated Contracting States: (74) Representative: Warner, James Alexander et al AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU Carpmaels & Ransford MC NL PT SE One Southampton Row London (30) Priority: 10.11.1997 US 64655 P WC1B 5HA (GB)

(43) Date of publication of application: (56) References cited: 03.09.2008 Bulletin 2008/36 • KWONG Y L ET AL: "Delicious poison: for the treatment of " BLOOD, (60) Divisional application: vol. 89, no. 9, 1 May 1997 (1997-05-01), pages 10177319.0 / 2 255 800 3487-3488, XP002195910 • SHEN S-X ET AL: "USE OF ARSENIC TRIOXIDE (62) Document number(s) of the earlier application(s) in (AS2O3) IN THE TREATMENT OF ACUTE accordance with Art. 76 EPC: PROMYELOCITIC LEUKEMIA (APL): II. CLINICAL 98957803.4 / 1 037 625 EFFICACY AND IN RELAPSED PATIENTS" BLOOD, W.B. (73) Proprietor: MEMORIAL SLOAN-KETTERING SAUNDERS, PHILADELPHIA, VA, US, vol. 89, no. CENTER 9, 1 May 1997 (1997-05-01), pages 3354-3360, New York, NY 10021 (US) XP000891715 ISSN: 0006-4971 • KONIG ET AL: "Comparative activity of (72) Inventors: melarsoprol and arsenic trioxide in chronic B - • Warrell, Raymond, P., Jr. cell leukemia lines" July 1997 (1997-07), BLOOD, Westfield, NJ 07090 (US) W.B. SAUNDERS, PHILADELPHIA, VA, US, VOL. • Pandolfi, Pier Paolo 90, NR. 2, PAGE(S) 562-570 , XP002095650 ISSN: New York, NY 10022 (US) 0006-4971 * page 5621, right-hand column, • Gabrilove, Janice L. paragraph 3 * New York, NY 10028 (US) • GABRILOVE ET AL: "Effects of arsenicals in chronic " ANTICANCER RESEARCH, HELENIC ANTICANCER INSTITUTE, ATHENS, vol. 17, no. 5C, 1 September 1997 (1997-09-01), page 3981, XP002095648 ISSN: 0250-7005

Note: Within nine months of the publication of the mention of the grant of the European patent in the European Patent Bulletin, any person may give notice to the European Patent Office of opposition to that patent, in accordance with the Implementing Regulations. Notice of opposition shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention). EP 1 964 557 B1

Printed by Jouve, 75001 PARIS (FR) EP 1 964 557 B1

Description

1. FIELD OF INVENTION

5 [0001] The present invention relates to a method for preparing a pharmaceutical composition which comprises arsenic trioxide.

2. BACKGROUND OF THE INVENTION

10 2.1. CANCER

[0002] Cancer is characterized primarily by an increase in the number of abnormal cells derived from a given normal tissue, invasion of adjacent tissues by these abnormal cells, and lymphatic or blood- borne spread of malignant cells to regional lymph nodes and to distant sites (). Clinical data and molecular biologic studies indicate that cancer 15 is a multistep process that begins with minor preneoplastic changes, which may under certain conditions progress to neoplasia. [0003] Pre-malignant abnormal cell growth as exemplified by hyperplasia, metaplasia, and dysplasia (for review of such abnormal growth conditions, see Robbins and Angell, 1976, Basic Pathology, 2d Ed., W.B. Saunders Co., Phila- delphia, pp. 68-79) precedes the formation of a neoplastic lesion. A neoplastic lesion may evolve clonally to grow into 20 a solid tumor, and develop an increasing capacity for invasion, growth, metastasis, and heterogeneity, especially under conditions in which the neoplastic cells escape the host’s immune surveillance (Roitt, I., Brostoff, J and Kale, D., 1993, Immunology, 3rd ed., Mosby, St. Louis, pps. 17.1-17.12). [0004] Leukemia refers to malignant of the blood-forming tissues. Transformation to malignancy typically occurs in a single cell through two or more steps with subsequent proliferation and clonal expansion. In some leukemias, 25 specific chromosomal translocations have been identified with consistent leukemic cell morphology and special clinical features (e.g., translocations of 9 and 22 in chronic myelocytic leukemia, and of 15 and 17 in acute promyelocytic leukemia). Acute leukemias are predominantly undifferentiated cell populations and chronic leukemias more mature cell forms. [0005] Acute leukemias are divided into lymphoblastic (ALL) and non- lymphoblastic (ANLL) types. They may be further 30 subdivided by their morphologic and cytochemical appearance according to the French- American-British (FAB) classi- fication or according to their type and degree of differentiation. The use of specific B- and T-cell and myeloid-antigen monoclonal antibodies are most helpful for classification. ALL is predominantly a childhood disease which is established by laboratory findings and bone marrow examination. ANLL, also known as acute myeloblastic leukemia (AML), occurs at all ages and is the more common acute leukemia among adults; it is the form usually associated with irradiation as 35 a causative agent. [0006] Chronic leukemias are described as being lymphocytic (CLL) or myelocytic (CML). CLL is characterized by the appearance of mature lymphocytes in blood, bone marrow, and lymphoid organs. The hallmark of CLL is sustained, absolute lymphocytosis (> 5,000/mL) and an increase of lymphocytes in the bone marrow. Most CLL patients also have clonal expansion of lymphocytes with B- cell characteristics. CLL is a disease of older persons. In CML, the characteristic 40 feature is the predominance of granulocytic cells of all stages of differentiation in blood, bone marrow, , spleen, and other organs. In the symptomatic patient at diagnosis the total WBC count is usually about 200,000/ mL, but may reach 1,000,000/mL. CML is relatively easy to diagnose because of the presence of the Philadelphia chromosome. [0007] The very nature of hematopoietic cancer necessitates using systemic as the primary treatment modality. Drugs selected according to sensitivities of specific leukemias are usually given in combination. Radiation 45 therapy may be used as an adjunct to treat local accumulations of leukemic cells. Surgery is rarely indicated as a primary treatment modality, but may be used in managing some complications. Bone marrow transplantation from an HLA- matched sibling is sometimes indicated.

2.2. ARSENIC AND ITS MEDICAL USES 50 [0008] Arsenic has been considered to be both a poison and a drug for a long time in both Western and Chinese medical practices. In the latter part of the nineteenth century, arsenic was used frequently in attempts to treat diseases of the blood in the West. In 1878, it was reported that treatment of a leukemic patient with Fowler’s solution (a solution containing , valence +5) reduced markedly the count of white blood cells (Cutler and Bradford, Am. 55 J. Med. Sci., January 1878, 81-84). Further interests in the use of Fowler’s solution as a palliative agent to treat chronic myelogenous leukemia (CML) was described by Forkner and Scott in 1931 (J. Am. Med. Assoc., 1931, iii, 97), and later confirmed by Stephens and Lawrence in 1936 (Ann. Intern. Med. 9, 1488-1502). However, while the active chemical ingredient(s) of Fowler’s solution was not determined, its toxicity was well recognized. Fowler’s solution was administered

2 EP 1 964 557 B1

strictly as an oral composition, and was given to leukemic patients as a solution until the level of white blood cells was depressed to an acceptable level or until toxicities (such as skin keratoses and hyperpigmentation) developed, while the patients enjoyed varying periods of remission. In the 1960’s, Fowler’s solution was still used occasionally in attempts to treat CML, however, most patients with CML were treated with other chemotherapeutic agents, such as , 5 and/or (Monfardini et al., Cancer, 1973, 31:492-501). [0009] Paradoxically, one of the long recognized effects of exposure to arsenic, whether the source is environmental or medicinal, is skin cancer (Hutchinson, 1888, Trans. Path. Soc. Lond., 39: 352; Neubauer, 1947, Br. J. Cancer, 1: 192). There were even epidemiological data to suggest that the use of Fowler’s solution over long periods could to an increased incidence of cancer at internal sites (Cuzick et al., Br. J. Cancer, 1982, 45: 904-911; Kaspar et al., J. Am. Med. 10 Assoc., 1984, 252: 3407-3408). The carcinogenicity of arsenic has since been demonstrated by the fact that it can induce chromosomal aberration, gene amplification, sister chromatid exchanges and cellular transformation (See e.g., Lee et al., 1988, Science, 241:79-81; and Germolec et al., Toxicol. Applied Pharmacol., 1996, 141:308-318). Because of the known carcinogenic effect of arsenic, its only therapeutic use in human in Western medicine today is in the treatment of tropical diseases, such as African trypanosomiasis, (the organic arsenical, melarsoprol; See Goodman & Gilman’s 15 The Pharmacological Basis of Therapeutics, 9th edition, chapter 66, 1659-1662, 1997). [0010] In traditional chinese medicine, or arsenic trioxide paste has been used to treat tooth marrow diseases, psoriasis, syphilis and rheumatosis (Chen et al., 1995, in Manual of Clinical Drugs, Shanghai, China, Shanghai Institute of Science and Technology, p.830). In 1970’s, arsenic trioxide had been applied experimentally to treat acute promyelocytic leukemia (APL) in China (commented by Mervis, 1996, Science, 273: 578). The clinical efficacy of arsenic 20 trioxide has recently been re- investigated in 14 of 15 patients with refractory APL, where the use of an intravenous dose at 10 mg/day for 4-9 weeks was reported to result in complete morphologic remission without associated bone marrow suppression (Shen et al., 1997, Blood, 89:3354-3360). It was also shown that arsenic trioxide induced apoptosis (pro- grammed cell death) in vitro in NB4 cells, an APL cell line, and that apoptosis was apparently associated with down- regulation of the oncogene bcl-2, and intracellular redistribution of the chimeric PML/RARα protein that are unique to 25 APL cells (Chen et al., 1996, Blood, 88:1052-1061; Andre et al., 1996, Exp. Cell Res. 229: 253-260). It has been reported that the biological activity of arsenic is due to the ability of arsenic to direct the nucleoplasmic fraction of PML to nuclear bodies for degradation (Zhu et al., 1997, Proc. Natl. Acad. Sci., 94: 3978-3983). [0011] Although arsenic is well known to be both a poison and a carcinogenic agent, there have been many reports concerning the use of arsenic in medical treatment. Further, from the above discussion, it should be clear that there are 30 many different types of leukemias, each of which requires a unique treatment protocol that is modified according to the presence of factors predicting for a risk of treatment failure. Thus, the development of a broad spectrum anti- leukemia agent that can be used alone or in combination with other existing drugs is extremely desirable.

3. SUMMARY OF THE INVENTION 35 [0012] Despite the conflicting reports in the art concerning benefits and risks of the administration of arsenic to patients, applicants have discovered that arsenic trioxide has broad applicability in the treatment of various types of leukemias, lymphomas, and solid tumors. [0013] The invention is directed to a method for the manufacture of pharmaceutical compositions comprising arsenic 40 trioxide. [0014] As described herein, arsenic trioxide can be used alone or in combination with other known therapeutic agents (including chemotherapeutics, radioprotectants and radiotherapeutics) or techniques to either improve the quality of life of the patient, or to treat leukemia, lymphoma or solid tumor. The arsenic compounds can be used before, during or after the administration of one or more known chemotherapeutic agents, including antitumor agents. In addition, the 45 arsenic compounds can be used before, during or after radiation treatment. [0015] The pharmaceutical compositions made by the method of the invention are sterile solutions suitable for intra- venous injection or infusion. [0016] Particular compositions which may be produced by the method of the invention and their uses are described in the sections and subsections which follow. 50 4. DETAILED DESCRIPTION OF THE INVENTION

[0017] Methods and compositions for the treatment of leukemia, lymphoma or solid tumors are described herein. In particular, a method of treating acute or chronic leukemia, lymphoma, or solid tumors in a human which comprises 55 administering to a human in need of such therapy a therapeutically effective and non-lethal amount of arsenic trioxide is described. Also described is a method of treating leukemia in a human who has become refractory to other forms of treatment which comprises administering to a human arsenic trioxide or melarsoprol in combination with another chem- otherapeutic agent, e.g., all-trans retinoic acid (ATRA).

3 EP 1 964 557 B1

[0018] The invention relates to a method for the manufacture of pharmaceutical compositions comprising arsenic trioxide. It is preferred that pharmaceutical compositions of the present invention exhibit reduced toxicity, improved efficacy, improved stability during storage and use, and that the composition has a physiologically acceptable pH.

5 4.1. THE ARSENIC COMPOUNDS

[0019] Since the non-pharmaceutically formulated raw materials used in the method of the invention are well known, they can be prepared from well-known chemical techniques in the art. (See for example, Kirk- Othmer, Encyclopedia of Chemical Technology 4th ed. volume 3 pps. 633-655 John Wiley & Sons). 10 [0020] As used herein the terms "a therapeutic agent", "therapeutic regimen", "radioprotectant", "chemotherapeutic" mean conventional drugs and drug therapies, including vaccines, for treating cancer, viral infections, and other malig- nancies, which are known to those skilled in the art. "Radiotherapeutic" agents are well known in the art. [0021] Arsenic trioxide or melarsoprol compounds can be used alone or in combination with other known therapeutic agents (including chemotherapeutics, radioprotectants and radiotherapeutics) or techniques to either improve the quality 15 of life of the patient, or to treat leukemia, lymphoma or solid tumor. For example, arsenic trioxide can be used before, during or after the administration of one or more known antitumor agents including but not limited to mustard compounds, mustard, , , , 6-, 6-thioguanine, , 5-fluor- ouracil, , , , , taxol, , temiposide, , , dox- orubicin, , mitomycin, , , , hydroxyurea, BCNU, , VM- 20 26, interferons, and all-trans retinoic acid (ATRA), or other (See, for example, the Physician Desk References 1997). In addition, arsenic trioxide can be used before, during or after radiation treatment.

4.2. METHODS OF TREATMENT

25 [0022] The term "a method for treating leukemia" as used herein means that the disease and the symptoms associated with the disease are alleviated, reduced, cured, or placed in a state of remission. For example, the methods of treatment of the invention can lower the white blood cell count, or reduce lymphocytosis in a human under treatment. [0023] The term "a method for treating lymphoma" as used herein means that the disease and the symptoms associated with the disease are alleviated, reduced, cured, or placed in a state of remission. 30 [0024] The term "a method for treating solid tumor" as used herein means that the disease and the symptoms associated with the solid tumor are alleviated, reduced, cured, or placed in a state of remission. [0025] In addition, the term "a method for treating leukemic infiltration" means that the infiltration of leukemic cells out of circulation and into other organs and systems and the symptoms associated with such infiltration are alleviated, reduced, cured, or placed in a state of remission. 35 [0026] The term "refractory" when used herein means that the leukemia is generally resistant to treatment or cure. [0027] As used herein, "preneoplastic" cell refers to a cell which is in transition from a normal to a neoplastic form; or cells that fail to differentiate normally; and morphological evidence, increasingly supported by molecular biologic studies, indicates that preneoplasia progresses through multiple steps. [0028] Arsenic trioxide (As 2O2) inhibits growth and induce apoptosis in NB4 acute promyelocytic leukemic cells. Acute 40 promyelocytic leukemia (APL) is associated with the t (15; 17) translocation, which generates a PML/ RARα fusion protein between PML, a growth suppressor localized on nuclear matrix-associated bodies, and RARα, a nuclear receptor for retinoic acid (RA). PML/RARα was proposed to block myeloid differentiation through inhibition of nuclear receptor re- sponse, as does a dominant negative RARα mutant. In addition, in APL cells, PML/RARα displaces PML and other nuclear body (NB) antigens onto nuclear microspeckles, likely resulting in the loss of PML and/or NB functions. It has 45 been suggested that high concentrations of arsenic trioxide promote apoptosis, whereas low concentrations induce partial differentiation in NB4 cells as well as cells derived from APL patients. It was postulated that As 2O3 works through its ability to specifically cause PML-RARα in APL cells to be relocalized to nuclear bodies for degradation (Zhu et al., 1997, Proc. Natl. Acad. Sci. USA, 94:3978-3983). However, these findings tend to limit the use of arsenic trioxide to a subset of leukemias. See Konig et al., 1997, Blood, 90:562-570. 50 [0029] As2O3 is able to inhibit cell growth, and induce apoptosis in various myeloid leukemia cell lines in a PML and PML-RARα independent manner and effective against a broad range of leukemias regardless of the underlying molecular mechanism that causes the neoplasia. Working examples of the effect of arsenic compounds on a number of leukemic cell lines are provided in Sections 5.1 and 5.2. [0030] Accordingly, the arsenic trioxide of the present invention can be used against a variety of leukemias, including 55 but not limited to:

Acute lymphoblastic leukemia (ALL) Acute lymphoblastic B-cell leukemia

4 EP 1 964 557 B1

Acute lymphoblastic T-cell leukemia Acute myeloblastic leukemia (AML) Acute promyelocytic leukemia (APL) Acute monoblastic leukemia 5 Acute erythroleukemic leukemia Acute megakaryoblastic leukemia Acute myelomonocytic leukemia Acute undifferentiated leukemia Chronic myelocytic leukemia (CML) 10 Chronic lymphocytic leukemia (CLL)

[0031] The skilled artisan will recognize that other leukemias may be treated in accordance with the present invention.

4.3. PROCESS FOR THE MANUFACTURE OF STERILE ARSENIC TRIOXIDE SOLUTION 15 [0032] The arsenic trioxide of the invention may be formulated into sterile pharmaceutical preparations for adminis- tration to humans for treatment of leukemias, lymphomas and solid tumors. Compositions comprising a compound of the invention formulated in a compatible pharmaceutical carrier may be prepared, packaged, labelled for treatment of and used for the treatment of the indicated leukemia, lymphoma, or solid tumor. 20 [0033] The invention provides a method for the manufacture of a pharmaceutical composition comprising a therapeutic effective and non-lethal amount of arsenic trioxide (As 2O3). Arsenic trioxide (raw material) is a solid inorganic compound that is commercially available in a very pure form. However, it is difficult to dissolve As 2O3 in aqueous solution. Further, the inventors are unaware of any published teachings on how to formulate As2O3 as a pharmaceutical composition suitable for injection directly into a human. Arsenic is present in solution in the +5 valence state (pentavalent) or the +3 25 valence state (trivalent). For example, (KAsO2; which is present in Fowler’s solution) and salts of arsenious acid contain pentavalent arsenic. It is known that one form of arsenic is more toxic than the other. (Goodman & Gilman’s The Pharmacological Basis of Therapeutics, 9th edition, chapter 66, 1660, 1997). A fresh solution of arsenic trioxide containing arsenic in the trivalent state will be gradually oxidized to pentavalent state if exposed to air for a

prolonged period, and as a result of the accumulation of pentavalent arsenic, the relative toxicity of a solution of As 2O3 30 will change over time. (Id.) Furthermore, it is observed that the total amount of arsenic in solution decreases over time. This loss of material is caused by the progressive conversion of arsenic in the solution into (AsH3) which is a gaseous compound at room temperature. This is particularly problematic in pharmaceutical applications if the concen- tration of an active ingredient in the injected material cannot be controlled. It is also undesirable to allow arsine to escape from the solution into the atmosphere because arsine is also toxic. 35 [0034] The inventors have experimented and successfully developed a method for formulating arsenic trioxide which overcomes the above-described problems of solubility and stability. The method comprises solubilizing solid high purity

As2O3 in an aqueous solution at high pH, such as pH greater than 12. For example, a 5 M solution of hydroxide may be used. To aid solubilization and obtain a clear and homogenous solution, mechanical stirring and/or gentle heating may be applied. A solution of As2O3 can also be obtained by dissolving the solid compound overnight. Typically, a 40 solution of 1 M As2O3 is obtained by this method. However, this solution is too basic to be useful as a pharmaceutical composition.

[0035] To adjust the pH of the As2O3 solution, the solution is first diluted in water, for example, to a concentration of about 1 mg/mL, pH 12. The As2O3 solution is then backtitrated with acid, such as, (1 M to 5 M HCl), with constant stirring until the pH is about 8.0 to 8.5. Highly concentrated HCl is not suitable as it causes precipitation 45 to occur in the As2O3 solution. The partially neutralized As2O3 solution may then be sterilized for example, by filtration (e.g., through a 0.22 mm filter), and stored in sterile vials. [0036] To make a pharmaceutical composition that can be directly injected into a subject, the composition must be sterile, standard techniques known to the skilled artisan for sterilization can be used. See, e.g., Remington’s Pharma-

ceutical Science. the pH of the partially neutralized As 2O3 solution is further adjusted to near physiological pH by dilution 50 (10-100 fold) with a 5% dextrose solution. For example, 10mL of a partially neutralized As 2O3 solution can be added to 500 mL of a 5% dextrose solution to yield a final pH of about 6.5 to 7.5. The method of the invention reduces the oxidation of arsenic in solution. Pharmaceutical compositions containing arsenic trioxide manufactured by the method of the invention show improved stability and long shelf life.

55 4.4. PHARMACEUTICAL COMPOSITION AND MODES OF ADMINISTRATION

[0037] The magnitude of a therapeutic dose of an arsenic compound in the acute or chronic management of leukemia will vary with the severity of the condition to be treated and the route of administration. The dose, and perhaps dose

5 EP 1 964 557 B1

frequency, will also vary according to the age, body weight, condition and response of the individual patient. In general, the daily dose ranges of arsenic trioxide for the conditions described herein are generally from about 0.05 to about 5 mg per kg body weight administered in divided doses administered parenterally or orally or topically. A preferred total daily dose is from about 2.5 to about 40 mg of arsenic trioxide. Preferably the arsenic trioxide formulation of the invention 5 is given daily for a maximum of 60 days, or until remission, followed by two to ten additional cycles, each lasting about 25 days in duration. For example, depending on the body weight of a patient with acute promyelocytic leukemia, a daily dose of arsenic trioxide greater than or less than 10 mg can be administered. Alternatively, following the weight- based dosing regimen, a therapeutic effect can be obtained with a daily dose of arsenic trioxide less than 10 mg. [0038] For treatment of solid tumor, a preferred dosing regimen involves intravenous infusion of about 0.1 to about 5 10 mg per kg body weight per day for 5 days. This five- day treatment protocol is repeated once per month until the tumor growth tumor is inhibited or when the tumor shows signs of regression. [0039] The effect of the therapy with arsenic trioxide or melarsoprol on development and progression of cancer can be monitored by any methods known in the art, including but not limited to determining: levels of tumor specific antigens and putative biomarkers, e.g., carcinoembryonic antigens (CEA), alpha- fetoprotein; and changes in morphology and/or 15 size using computed tomographic scan and/or sonogram. [0040] Desirable blood levels may be maintained by a continuous infusion of an arsenic compound as ascertained by plasma levels. It should be noted that the attending physician would know how to and when to terminate, interrupt or adjust therapy to lower dosage due to toxicity, or bone marrow, liver or dysfunctions. Conversely, the attending physician would also know how to and when to adjust treatment to higher levels if the clinical response is not adequate 20 (precluding toxic side effects). [0041] The pharmaceutical compositions of the present invention comprise arsenic trioxide as the active ingredient, and a pharmaceutically acceptable carrier specifically 5% dextrose solution, and optionally, other therapeutic ingredients, for example all trans retinoic acid. [0042] In the case where an intravenous injection or infusion composition is employed, a suitable dosage range for 25 use is, e.g., from about one to about 40 mg arsenic trioxide total daily; about 0.001 to about 10 mg arsenic trioxide per kg body weight total daily, or about 0.1 to about 10 mg melarsoprol per kg body weight total daily. [0043] An exemplary course of treatment of a patient with leukemia, lymphoma, or solid cancer can involve daily administration by intravenous infusion of arsenic trioxide in an aqueous solution at a daily dose of about 0.01 to 1 mg arsenic trioxide per kg of body weight of the patient. Preferably, about 0.15 mg arsenic trioxide per kg body weight per 30 day is used. The course of treatment may continue until bone marrow remission is observed or when side effects are becoming serious. The course of treatment may be repeated for up to ten times over approximately 10 months with a break of about 3 to 6 weeks in between courses. The post-remission course of treatment involves infusion of arsenic trioxide at a daily dose of about 0.15 mg per kg of body weight of the patient on a daily or weekdays-only basis for a cumulative total of 25 days. 35 5. EXAMPLES

[0044] Described below are examples of uses of the arsenic compounds of the invention in treatment of various types of leukemia. Through these and other experiments the arsenic trioxide formulation of the invention were found to be 40 well-tolerated in humans. For example, three APL patients were given 10 mg of the arsenic trioxide formulation of the invention once daily (flat dose) intravenous dose.

5.1. ARSENIC TRIOXIDE AND MELARSOPROL INDUCE APOPTOSIS IN MYELOID LEUKEMIA CELL LINES

45 [0045] The activity of As2O3 and melarsoprol against myeloid leukemia cell lines, including the APL cell line NB4-306 (a retinoic acid-resistant cell line derived from NB4 that no longer expresses the intact PML- RARα fusion protein), HL60, KG-1, and the myelomonocytic cell line U937 was investigated. To examine the role of PML in mediating arsenical activity, the inventors also tested these agents using murine embryonic fibroblasts (MEFs) and bone marrow (BM) progenitors in which the PML gene had been inactivated by homologous recombination. Unexpectedly, it is found that 50 both compounds inhibited cell growth and induced apoptosis in all cell lines tested. Melarsoprol was more potent than -7 -5 As2O3 at equimolar concentrations ranging from 10 to 10 mol/L. As 2O3 relocalized PML and PML- RARα onto nuclear bodies, which was followed by PML degradation in NB4 as well as in HL60 and U937 cell lines. Although melarsoprol was more potent in inhibiting growth and inducing apoptosis, it did not affect PML and/or PML- RARα nuclear localization.

Moreover, both As 2O3 and melarsoprol comparably inhibited growth and induced apoptosis of PML+/ + and PML-/-MEF, 55 and inhibited colony-forming unit erythroid (CFU-E) and CFU granulocyte-monocyte formation in BM cultures of PML+/ + and PML-/- progenitors. A detailed description of the methods, materials, and results of these experiments is provided in Wang et al., Blood, 1998, 92:1497-1504. [0046] Results from the experiments show that the cytotoxic effect of both arsenicals in these cell lines is not mediated

6 EP 1 964 557 B1

by mechanisms that are dependent on PML or PML- RARα expression. In most lines, melarsoprol was somewhat more potent compared with As2O3 in inhibiting growth and inducing apoptosis, and the effects of both drugs were dose dependent. As previously reported, it is confirmed that As 2O3 relocalized PML protein onto nuclear bodies and induced PML and PML-RARα degradation in NB4 cells while triggering spoptosis. However, similar effects were also observed 5 in HL60 and U937 cells which do not harbor the PML-RARα fusion gene. Moreover, melarsoprol induced apoptosis in all the cell lines tested without altering PML and/or PML- RARα.

[0047] The differentiating action of As 2O3 and melarsoprol, appeared negligible in vitro, and did not appear to depend on the expression and/or modulation of PML and/or PML- RARα either. In fact, the small effect observed by the inventors in long-term cultures (up to 2 weeks), was comparable in all the cell lines tested with both compounds. 10 [0048] It is also found that bcl- 2 downregulation, which has been previously linked to the antileukemic effects of As 2O3 in APL, was also not dependent on expression of PML-RARα protein, because it occurred in the NB4 subclone 306 in which the intact protein is not detectable. Finally, to test whether PML expression was essential to the antileukemic effects of arsenicals, both agents were tested in mouse embryonic fibroblasts and BM cells from animals wherein wild-

type PML had been eliminated by homologous recombination. In these cells wholly lacking PML expression, both As 2O3 15 and melarsoprol were equally effective in inhibiting growth and inducing apoptosis, and both had similar effects on normal CFU-E and CFU-GM colony formation. Moreover, no differences between wild-type and PML-/- cells were observed. Without being limited by any theory, together, these data strongly support theory that the antileukemic effects of these arsenicals occurs independently of both PML and PML- RARα expression. These results are in keeping with the medicinal history of arsenicals for diseases that are not characterized by alterations in PML protein such as, for instance, chronic 20 myelocytic leukemia. [0049] The results indicate that both As 2O3 and melarsoprol are broadly active as antileukemic agents in both myeloid and lymphoid diseases. In conclusion, the data indicate that cytotoxic activity is not mediated by the PML protein and therefore is not limited to diseases that are associated with alterations in PML expression. Thus, the arsenic compounds of the invention have a potentially broader therapeutic role that is not confined to APL. 25 5.2 CLINICAL STUDY OF MELARSOPROL IN PATIENTS WITH ADVANCED LEUKEMIA

[0050] Melarsoprol, an organic arsenical synthesized by complexing melarsen with dimercaprol, has primarily been used for the treatment of African trypanosomiasis. The effects of melarsoprol upon induction of apoptosis in cell 30 lines representative of chronic B- cell lymphoproliferative disorders have been investigated, and the results are described below. [0051] Melarsoprol (supplied as Arsobal [36 mg/mL] by Rhone Poulenc Rorer, Collegeville, PA) was diluted in propylene -4 glycol at a stock concentration of 10 mol/L and stored at room temperature. As 2O3 (Sigma, St. Louis, MO) was dissolved in 1.65 mol/L sodium hydroxide (NaOH) at a stock solution of 10-3 mol/L. Serial dilution (10-6 to 10-7 mol/L were made 35 in RPMI 1640 media. An Epstein- Barr virus (EBV) - transformed B- prolymphocytic cell line (JVM- 2), an EBV-transformed B-cell chronic lymphocytic leukaemia (B-CLL) cell line (183CLL), and one non- EBV-transformed B-CLL cell line (WSU- CLL) were used as targets. Dose-response experiments with melarsoprol (10-7 to 10-9 mol/L) were performed over 96 hours. [0052] Unexpectedly, the inventors found that melarsoprol caused a dose- and time-dependent inhibition of survival 40 and growth in all three cell lines. In contrast. As 2O2 at similar concentrations had no effect on either viability or growth. After 24 hours, all three cell lines treated with melarsoprol (10 -7 mol/L) exhibited morphologic characteristics of apoptosis. A prominent concentration-dependent downregulation of bcl-2 mRNA after 24 hours of exposure to melarsoprol in WSU- CLL 183CLL, and JVM-2 cells was observed. Decrease of bcl-2 protein expression was also observed in all three cell lines, whereas As2O2 had no effect on this parameter. 45 [0053] Given that the in vitro data above have shown unexpectedly broad antileukemic activity for melarsoprol against both myeloid and lymphoid cells, and generally at lower concentrations than As2O3, a study was initiated to evaluate the pharmacokinetics, safety, and potential efficacy of melarsoprol in human patients with relapsed or refractory leukemia. [0054] Eligible patients were treated with a brief IV injection daily for 3 days, repeated weekly for 3 weeks, with an additional 3 wk course in responding pts. The initial dose was 1 mg/kg on Day 1, 2 mg/kg on Day 2, and 3.6 mg/kg on 50 Day 3 and all days thereafter. Parallel in vitro studies included culture sensitivity of fresh leukemic cells to both melarsopr ol and As2O3, along with serial flow cytometric studies of surface antigen expression, apoptosis, andbcl- 2 expression. Three patients with AML and one with CML have entered the study. [0055] Using a method based on high performance liquid chromatography that is sensitive to approximately 10 mg/ml, preliminary pharmacokinetic data show that peak plasma drug concentrations were obtained immediately after injection 55 with a Cmax that ranged from 1.2 ng/ml on day 1 to 2.4 ng/ml on day 3. While the initial distribution phase was rapid, a prolonged T©γ has suggested release from a deep compartment. Plasma areas under the concentration x time curves (AUCs) were proportional to the administered dose, ranging from 0.48 ng·hr/ml on Day 1 to 1.48 ng·hr/ml on Day 3. Detectable concentrations of the drug were found in plasma one week after initial dosing. The drug has been relatively

7 EP 1 964 557 B1

well-tolerated. Adverse effects have included transient pain at the injection site and mild . No signs of "reactive encephalopathy" (occasionally observed during treatment of CNS trypanosomiasis) have been observed. [0056] Results from these studies suggest that melarsoprol may have broader activity than inorganic As 2O3, and that concentrations which are cytotoxic to leukemic cells in vitro, and thus therapeutic, are readily achieved in vivo. 5 5.3. ARSENIC TRIOXIDE INDUCES APOPTOSIS IN K562 CHRONIC MYELOGENOUS LEUKEMIA (CML) CELLS

[0057] A Philadelphia chromosome positive CML cell line K562 is used to determine if arsenic trioxide (As 2O3) promotes -5 apoptosis in CML. Suspension cultures of cells in log phase were exposed to As2O3 at concentrations of 1 x 10 M, 5 10 x 10-6M, and 1 x 10-6M. Aliquots of cells were analyzed at various time points over the course of 72 hours to assess viability and apoptosis. Viability was measured using trypan blue exclusion; at the same time, apoptosis was detected by morphology, , and DNA gel electrophoresis. [0058] Arsenic trioxide at a concentration of 1 x 10-6M had no effect on K562 cell growth or viability. The greatest -5 effect on cell growth and survival was seen with 1 x 10 M As2O3. K562 cell growth and viability data after 72 hours of 15 exposure to As2O3 are recorded in Table 1:

Table 1: % Cell Growth Impairment % Viability p value Control 0 92.1 6 0.9 20 -6 5 x 10 M As 2O3 63.0 75.8 6 0.5 0.0001 -5 1 x 10 M As 2O3 75.3 61.9 6 2.9 0.0223

Evidence that this arsenic-induced decrease in viability represented apoptosis was analyzed. Morphologic features of 25 apoptosis including membrane blebbing and nuclear condensation were evident in stained cytospins of K562 cells -5 incubated with 10 M As2O3 for 72 hours. This correlated with evidence of DNA internucleosomal damage as visualized -5 by gel electrophoresis of DNA extracted from K562 cells exposed to 10 M As2O3. Quantitative assessment of apoptosis, -5 as measured by the TUNEL method demonstrated that 75.6% 6 8.6 (1 x 10 M As2O3) cells exhibited apoptosis as -5 comparedwith 6.3% 6 3.0 (control) cells at 72 hours. Treatment of K562 cells with 10 M As 2O3 resulted in an upregulation 30 of p21 mRNA, as detected by Northern analysis, suggesting an arrest of the cells in the G1 phase of the . This data indicates arsenic trioxide as a therapeutic agent for CML.

5.4. THERAPEUTIC TRIALS WITH RETINOIC ACID AND ARSENIC TRIOXIDE (As 2O3) IN PML-RARα AND PLZF- RARα TRANSGENIC MICE 35 [0059] Acute promyelocytic leukemia (APL) is associated with chromosomal translocations which invariably involve the translocation of the Retinoic Acid Receptor α (RARα) locus on chromosome 17 to other loci in the genome, such as in the majority of APL cases, the PML gene located on chromosome 15, and in a few cases the PLZF gene on chromosome 11. Patients harboring the t(15;17) are sensitive to treatment with All-Trans Retinoic Acid (ATRA), yielding complete 40 remission rates of 75% to 95%. APL associated with the t (11;17) (PLZF-RARα) shows a poor response to ATRA.

[0060] To test the efficacy of As2O3 in the treatment of APL, models of the disease were created in transgenic mice. Transgenic mice were generated by standard techniques in which the expression of the PML-RARα or PLZF-RARα fusion proteins is placed under the control of a myeloid-promyelocytic specific human Cathepsin-G (hCG) minigene. Both hCG-PML/RARα and hCG-PLZF-RARα transgenic mice develop myeloid leukemia with features of APL similar to 45 those in humans. [0061] Therapeutic trials on these leukemic mice with the following regimens were started: 1) ATRA: 1.5 mg per gram of body weight per day administered orally; and 2) ATRA: 6 mg per gram of body weight per day administered intraperi- toneally. Mice were bled once a week to evaluate the response. [0062] PML/RARα leukemias responded well to ATRA with high remission rates (80% with regimen 1). Surprisingly, 50 in vitro, ATRA induced differentiation, and inhibited growth of leukemic cells as well as leukemic colony formation in bone marrow and spleen progenitors assays in both PML-RARα and PLZF-RARα leukemias. Furthermore, in ex vivo experiments, leukemic cells from PLZF-RARα mice lost their tumorigenic capacity when transplanted in recipient nude mice upon preincubation with ATRA, while untreated cells were tumorigenic. However, in vivo, PLZF-RARα leukemias responded poorly to ATRA (28% with regimen 1), while higher doses of ATRA appeared more effective (50% with 55 regimen 2). In conclusion, leukemias in PLZF- RARα transgenic mice are sensitive to ATRA treatment, but might require therapeutic regimens with high doses of ATRA. These findings have direct implications in the treatment of APL patients with t(11;17). [0063] In both PML-RARα and PLZF- RARα leukemias, ATRA prolonged survival, but leukemias relapsed shortly after

8 EP 1 964 557 B1

remission was achieved, and were refractory to further ATRA treatment. The two transgenic mouse models is also used to test the efficacy and dosage of As2O3, and ATRA+As2O3 in combination for treatment of APL patients resistant to ATRA, and in APL associated with the t (11;17). A regimen of As2O3 at 6 mg per day or a combination of As 2O3 at 6 mg and ATRA at 1.5 or 6 mg per gram of body weight per day is administered intraperitoneally. Mice are bled weekly to 5 evaluate the remission of the APL.

5.5. MANUFACTURE AND STABILITY OF PHARMACEUTICAL FORMULATION

[0064] Solid ultrapure arsenic trioxide (As2O3) was solubilized in a solution of 5 M sodium hydroxide (NaOH). The 10 suspension was stirred at ambient temperature for 5 minutes which yielded a clear, homogenous solution. The As2O3 solution (2 mL, 1.0 M) was added to 393.6 mL of H 3O in a 500 ml Erlenmeyer flask, which yielded an As 2O3 concentration of 1 mg/mL at pH = 12. A 5.0 M HCl solution was prepared by dilution of HCl (49.26 mL, 37% wt/wt, 10/15 M) with H 2O (50.74 mL) in a 250 mL Erlenmeyer flask. The HCl solution was later transferred via syringe to a 1000 mL empty evacuated

container. The As2O3 solution was back titrated with HCl (0.725 mL, 5.0 M) to pH 8.0. Approximately 10 mL of the 15 backtitrated As2O3 solution was filtered through a Millex-GS 0.22 mm filter unit and was added to each of approximately 30 sterile evacuated sterile vials. To make the pharmaceutical composition which would be injected intravenously into patients, 10 mL of this solution was withdrawn from two of the vials and was added to a 500 mL 5%-dextrose solution which yielded a final pH of 6.5. [0065] The high purity of the bulk starting material was confirmed (see Table 1) by atomic absorptiometry. Duplicate 20 samples of four intermediate or final- step solutions were assayed for total arsenic content. Assay bulk powder confirmed the extremely high purity of the starting material. Data for arsenic content of the intermediate and finished product solutions are presented in Table 2 below. [0066] The data below show that the solutions are stable in that there does not appear to be any indication of weight loss of arsenic over time. 25 Table 2 Arsenic content (ppm) of intermediate formulation and finished product solution of arsenic trioxide. Sample Code A-01* A-02 A-03 A-04 A-05 Aliquot A 140,600 600 707 629 680 30 Aliquot B 139,000 564 703 688 687 Assay Variance 1.1% 6% 0.57% 8.7% * Identity of sample codes: A-01: Intermediate product solution after initial solubilization in NaOH. 35 A-02: Intermediate product solution prior to HCl titration. A-03: Intermediate product prior to Millex filtration. A-04: Finished product from sterile 10 ml fill vial immediately after manufacturing. A-05: Finished product from capped vials two months after manufacturing.

40 6. EXAMPLES: CLINICAL TRIALS IN APL PATIENTS

[0067] Arsenic trioxide was evaluated in patients with APL to determine whether this agent induced either cytodiffer- entiation or apoptosis. Twelve patients who had relapsed from extensive prior therapy were treated with arsenic trioxide

45 at doses ranging from 0.06 to 0.2 mg/kg per day until a bone marrow remission was achieved. Bone marrow mononuclear cells were serially monitored by flow cytometry for immunophenotype, fluorescence in situ hybridization (FISH), reverse polymerase chain reaction (RT-PCR) assay for PML/RAR-α expression, and Western blot expression of the apoptosis-associated proteins, caspases 1, 2 and 3. The results showed that low- doses of arsenic trioxide are highly effective for inducing complete remission in relapsed patients with APL. Clinical response is associated with incomplete

50 cytodifferentiation and induction of apoptosis with caspase activation in leukemic cells.

6.1. METHODS

[0068] Clinical protocol: Eligibility requirements included a diagnosis of APL confirmed by cytogenetics or fluorescence

55 in situ hybridization (FISH) analysis for a t(15;17) translocation, or by reverse transcriptase polymerase reaction (RT- PCR) assay for PML/RAR- α. Patients must have relapsed from standard therapy that had included all- trans retinoic acid plus a combination of cytotoxic drugs. Signed informed consent was required, and the protocol was reviewed and approved by this center’s institutional review board

9 EP 1 964 557 B1

[0069] Arsenic trioxide treatment: Arsenic trioxide was supplied as an aqueous solution in 10 ml vials containing 1 mg/ml of drug. The drug was further diluted in 500 ml of 5%-dextrose solution and infused intravenously over 2 to 4 hours once per day. While the initial cohort of patients received either 10 or 15 mg/day as a flat dose, the referral of two children prompted the invention of a weight-based regimen (0.15 mg/kg/day) that was heretofore unknown. The drug 5 was given daily until bone marrow remission was observed. Patients who achieved complete remission were eligible for treatment with additional courses of therapy 3 to 6 weeks after the preceding course. Subsequent courses were generally given at a dose of 0.15 mg/kg/day for a cumulative total of 25 days, administered either daily or on a weekdays-only schedule, for a maximum total of 6 courses over approximately 10 months. Monitoring during study: Patients with coagulopathy were transfused with and fresh-frozen plasma to maintain the count and fibrinogen at 10 target levels ≥ 50,000 cells/cu mm and ≥ 100 mg/dL, respectively. Blood counts, coagulation studies, serum chemistry profiles, urinalyses, and electrocardiograms were serially obtained. A bone marrow aspiration and/or biopsy was per- formed at baseline and periodically thereafter until remission was documented. Conventional response criteria were observed, which included recovery of bone marrow to ≤ 5% blasts, peripheral blood leucocytes ≥ 3,000 cells/cu mm, and platelets ≥ 100,000 cells/cu mm. 15 [0070] Cellular immunophenotype studies: Heparinized bone marrow or blood samples were collected and mononu- clear cells were isolated by Ficoll- Hypaque centrifugation. Surface membrane antigens were detected by direct immun- ofluorescence staining using fluorescein isethiocynate (FITC) or phycoerythrin conjugated monoclonal antibodies: CD16 (Leu 11a), CD11b, CD33 (Leu M9), HLA- DR, CD45, and CD14, purchased from either Becton- Dickinson (Mountainview, CA) or Immunotech Immunology (Marseille, France). Dual-color staining was performed by incubating cells simultane- 20 ously with two monoclonal antibodies, including CD33-PE/CD11b-FITC and CD33-PE/CD16-FITC. Negative controls using irrelevant monoclonal immunoglobulins of the same isotype were analyzed concurrently. Flow cytometric analyses were performed on an EPICS Profile II flow cytometer (Coulter Electronics) equipped with a 488 nm argon laser. Forward and side-scatter cell parameters were measured and combined with CD45/CD14 staining to identify populations of interest and to exclude monocytes from the analysis gate. The Multiparameter Data Acquisition and Display System 25 (MDADS, Coulter Electronics) was used to acquire and analyze data. [0071] Fluorescence in situ hybridization (FISH): Selected specimens that had undergone immunofluorescence stain- ing for CD33 and CD11b were sorted for cells that coexpressed both antigens using a FACStar Plus cell- sorter (Becton- Dickinson). Separated cells were incubated in culture media at 37° C for one hour, treated with hypotonic solution 0.075M KCl for 5 minutes, fixed in 3:1 methanol: fixative, and air-dried. Interphase FISH was performed using a 30 specific PML/RAR-α translocation dual-color probe (Vysis; Downer’s Grove, IL). Briefly, DNA from interphase cells was denatured by immersing slides in a solution of 50% formamide/2xSSC at 73° C for 5 minutes; the slides were then dehydrated in alcohol and air dried. A mixture of probe in hybridization mixture was applied, covered with a cover slip, and sealed with rubber cement. Hybridization was carried out at 37° C in a moist chamber for approximately 12 to 16 hours. Following hybridization, unbound probe was removed by washing the slides at 45° C in 50% formamide/ 2xSSC 35 solution three times for 10 minutes each, followed by a wash in 2xSSC/ 0.1 NP-40 solution at 45° C for 5 minutes. Slides were then air dried and counter- stained with 4’,6-diamidino-2-phenylindole and covered with a glass coverslip. Analysis of interphase cells for fluorescent signals was performed with a Photometrics Sensys camera fitted to a Zeiss axioscope. A minimum of 300 cells was studied for each sample. [0072] Western blot analysis: Cells were lysed in a buffer containing 50 mM Tris- HCl, 0.5 mM ethylene glycol [bis]-[amin- 40 ioacyl] tetra acetic acid, 170 mM NaCl, 1mM dithiothreitol, 0.2% NP-40, 0.01 U/mL aprotinin, 10m g/mL leupeptin, 10mg/mL , and 1 mM phenylmethylsulfonyl fluoride (all from Sigma). The lysates were then sonicated using a ultrasonic homogenizer (471C series, Cole Parmer Instruments, Chicago, IL) and centrifuged at 7,500 g (Sorvall Instru- ments, Newtown, CT). Protein content of the lysates was determined using a BioRad Protein Assay Kit (Bio-Rad Lab- oratories, Hercules, CA) at 595 nm with a BSA standard. A sample buffer containing 10 % glycerol, 0.4 % SDS, 0.3 % 45 bromphenol blue, 0.2 % pyronin Y, in 1x stacking buffer (Tris base 0.5 M, 0.8 % SDS), 20 % 2-mercaptoethanol, was added to the cell lysates, which were heat- denatured at 95°C for 3 min. Subsequently, 15 mg/lane of protein was loaded on a SDS-polyacrylamide gel containing 12.5% polyacrylamide and was size-fractionated by electrophoresis. Proteins were electroblotted onto Trans-Blot® transfer medium (Bio-Rad) and stained with Ponceau-S as an internal loading control. Rabbit anti- human monoclonal antibodies, including caspase 1, caspase 2 (both from Santa Cruz Biotechnology, 50 Santa Cruz, CA), and caspase 3 (PharMingen, San Diego, CA) were added, and bound antibodies were detected using the ECL™ chemiluminescence detection system (Amersham, Arlington Heights, IL). Protein bands were quantified by computer densitometry. [0073] RT-PCR analysis for PML/RAR-α: RT-PCR was performed using methods previously described (Miller et al., 1992, Proc. Natl. Acad. Sci. 89:2694-8; Miller et al., 1993, Blood, 82:1689-94). 55 6.2. RESULTS

[0074] Patients: Twelve patients with relapsed or refractory APL were treated. All patients had received extensive

10 EP 1 964 557 B1

prior therapy with retinoids and cytotoxic drugs (Table 3). Two patients had relapsed from allogeneic bone marrow transplantation, one of whom had also failed donor T- cell reinfusion. One patient was being maintained on hemodialysis for chronic renal failure. Clinical Efficacy: Eleven of the 12 patients achieved a complete remission after arsenic trioxide treatment. The patient 5 who entered on hemodialysis sustained an intracranial hemorrhage on day 1 and died on day 5. The median duration of therapy in responding patients was 33 days (range, 12 to 39 days), the median daily dose was 0.16 mg/kg (range, 0.06 to 0.2 mg/kg), and the median cumulative dose during induction was 360 mg (range, 160 to 515 mg) (Table 3). Complete remission by all criteria was attained at a median time of 47 days (range, 24 to 83 days) after initiation of therapy. Remission by bone marrow criteria - the determining factor for discontinuing therapy - was achieved first, usually 10 followed in sequence by recovery of peripheral blood leukocytes and platelets. Over the range of doses used in this study, no differences in efficacy or time to response were obvious. After 2 courses of therapy, 8 of 11 patients tested had converted their RT-PCR assays for PML/RAR-α from positive to negative. [0075] All 11 patients in complete remission completed at least 1 post- remission treatment course with arsenic trioxide. Four, two, and one patient each have completed a total of three, four and five treatment courses, respectively. The 15 median duration of remission is 5+ months (range, 1 to 9+ months). However, 3 of the 11 patients relapsed during the second treatment course; none of these patients had converted their RT-PCR assays, and each appeared to have rapidly acquired drug resistance. Two of these individuals have since expired from progressive leukemia. [0076] Adverse Events: The clinical condition of patients in this study was highly variable, which reflected their extensive prior therapy. The protocol did not require hospitalization; three patients completed induction therapy entirely as outpa- 20 tients, and one other individual was hospitalized solely for placement of a venous catheter. However, 8 patients were hospitalized for complications of leukemia, 5 of whom required transfer to an intensive care unit, endotracheal intubation, and assisted ventilation for complications that included pulmonary hemorrhage, renal failure, sepsis, graft vs. host disease, non-specific pulmonary infiltrates, or hypotension. One patient required insertion of a permanent pacemaker after second-degree heart block developed in the setting of severe metabolic acidosis, hyperkalemia, hypotension, and 25 renal insufficiency. However, the heart block reversed despite rechallenge with further arsenic trioxide therapy. The drug was temporarily withheld due to serious intercurrent medical complications in 5 patients for a median of 2 days (range, 1 to 5 days). Two patients developed symptoms similar to that of the "retinoic acid syndrome"; both were presumptively treated with dexamethasone and improved. Only 2 patients required no platelet transfusions whatsoever; the median number of platelet units transfused was 61 (range, 0 to 586 units). 30 [0077] The median total peripheral blood leukocyte count at entry was 4,700 cells/cu mm (range, 500 to 144,000 cells/cu mm). Six patients developed leukocytosis (i.e., ≥ 20,000 cells/cu mm) that ranged from 20,800 to 144,200 cells/cu mm. No additional therapy was administered to these patients, and the leukocytosis resolved in all cases without further intervention. [0078] Common adverse reactions included lightheadedness during the infusion, , musculoskeletal pain, and 35 mild hyperglycemia. Three patients developed dysesthias presumably due to peripheral neuropathy. However, 2 of these patients had been immobilized for prolonged periods during assisted ventilation, and the other patient had an antecedent neuropathic history. [0079] Immunophenotype studies: APL is characterized by cells that express CD33, an antigen typically associated with primitive myeloid cells. Arsenic trioxide therapy induced a progressive decrease in the proportion of cells that solely 40 expressed CD33, along with an increase in the proportion of cells that expressed CD11b, an antigen associated with mature myeloid elements. While these changes would be anticipated from any agent that induced remission in APL, arsenic trioxide also induced expression of cells that simultaneously expressed both antigens. In most cases, these dual-expressing cells dominated the myeloid cell population, and they persisted for extended periods after complete remission was achieved by clinical criteria. 45 [0080] Fluorescence in situ hybridization analysis: Bone marrow mononuclear cells taken from a patient both early and later in complete remission were sorted by flow cytometry for coexpression of CD33 and CD11b. Using fluorescence in situ hybridization (FISH) analysis, three hundred cells were examined early in remission. Similar to control APL cells, the majority of these cells yielded a hybrid signal, indicating a translocation between PML and RAR-α genes and their origin from the neoplastic clone. However, when cells from the same patient were again sorted using these same 50 parameters later in remission, only the normal pattern of fluorescence signals was detected, indicating their derivation from normal hematopoietic progenitors. [0081] Western blot analysis: Protein extracts from bone marrow mononuclear cells were serially examined by Western blot analysis. The analysis showed that the precursor forms of caspase 2 and caspase 3 were upregulated in vivo in response to arsenic trioxide treatment. Moreover, this treatment also induced expression of cleaved fragments of caspase 55 1, indicating activation of the . There is also some indication that expression of the cleaved form of caspase 3 is increased. The antibody used in these experiments does not react with the cleaved form of caspase 2.

11 EP 1 964 557 B1

6.3. DISCUSSION

[0082] In this study, with few exceptions, patients admitted to the trial had sustained multiple relapses and were resistant to conventional chemotherapy, retinoids, or bone marrow transplantation. At entry, patients in this study suffered 5 from numerous leukemia-related complications, including respiratory failure, disseminated Varicella zoster infection, cavitary aspergillosis, chronic renal failure, and graft- vs.-host disease. Moreover, 5 of the 12 patients required admission to an intensive care unit for assisted ventilation and supportive care, but these complications were not directly related to arsenic trioxide therapy. [0083] Virtually all patients with a confirmed diagnosis of APL attained remission without the early mortality associated 10 with therapy. Although less commonly observed compared with all- trans retinoic acid treatment, arsenic trioxide induced striking leukocytosis in several patients. Upon withholding other cytotoxic drugs, the leukocytosis disappeared as patients attained remission. Despite 3 early relapses, 8 of 11 patients tested converted RT- PCR assays for PML/RAR- α (a molecular marker of residual disease) to negative, a phenomenon that is unusual after all- trans retinoic acid treatment alone. Finally, arsenic trioxide is active in APL over at least a three- fold dose range from 0.06 to 0.20 mg/kg. 15 [0084] All-trans retinoic acid induces "terminal" differentiation of APL cells, but the cytodifferentiating effects of arsenic trioxide appear to be incomplete. Arsenic induces a population of cells that simultaneously express surface antigens characteristic of both mature and immature cells (i.e. CD11b and CD33, respectively). Early during induction, these cells retain the t(15;17) translocation that characterizes APL. Unexpectedly, these cells persisted in the bone marrow despite the achievement of a clinically complete remission; however, later in remission, the coexpressing cells - while still readily 20 detectable - were no longer positive by in situ hybridization. The morphologic appearance of leukemic cells during therapy is also far less distinctive than that observed during therapy with all- trans retinoic acid. In fact, leukemic cells from many patients displayed few morphologicchanges for 10or more days, after whichthe proportionof leukemic cellsprogressively decreased. [0085] Following "non- terminal" differentiation, arsenic trioxide appeared to induce apoptosis, coincident with increased 25 expression and conversion of cysteine proteases (termed caspases) from inactive precursors to activated . The caspase pathway has only recently been characterized as an important pathway of programmed cell death. Initially recognized due to homology between the C. elegans protein ced-3 and mammalian interleukin-1β converting enzyme (ICE), the family of caspases now encompasses at least 10 different proteins that cleave a number of polypeptides. In leukemic cell lines, caspase activation is inducible with a number of cytotoxic agents, including all-trans retinoic acid. 30 Since these enzymes induce widespread proteolysis, it is conceivable that PML/RAR- α is a caspase substrate. [0086] A final similarity shared by arsenic trioxide and all- trans retinoic acid is the rapid development of clinical resist- ance in some individuals. Leukemic cells taken from two patients who relapsed retained in vitro sensitivity over concen- trations ranging from 10-4M to 10-7M. Relative arsenic resistance due to decreased intracellular transport has been described in association with down-regulation of membrane transporters encoded by the ars operon in bacterial cells. 35 Resistance in mammalian cells is less well-characterized, but alterations in membrane transport or efflux are probably important factors. [0087] In summary, arsenic trioxide induces complete remission in patients with APL who have relapsed from extensive prior therapy. This drug causes partial but incomplete cytodifferentiation of leukemic cells, followed by caspase activation and induction of apoptosis. 40 Table 3: Clinical characteristics and induction therapy results of patients with acute promyelocytic leukaemia treated with arsenic trioxide. Age No. of Treatment Daily Cumulative Time to Platelets ≥ Leukocytes ≥ (yrs) Relapses duration Dose Dose (mg) Remission 100,000/cu 3,000/cu Mm 45 (days) (mg/kg) (days) Mm 36 1* 36 .16 360 54 36 54 45 3*a 39 .12 390 83 39 83 a,b 50 31 3 37 .18 370 41 39 41 25 2 16 .06 160 24 16 16 62 2*d 30 .11 300 41 41 31 75 1 12 .20 180 30 30 30 55 40 1 * 33 .16 495 47 47 43 13 2*a,b 27 .18 270 50 41 52

12 EP 1 964 557 B1

(continued) Age No. of Treatment Daily Cumulative Time to Platelets ≥ Leukocytes ≥ (yrs) Relapses duration Dose Dose (mg) Remission 100,000/cu 3,000/cu Mm (days) (mg/kg) (days) Mm 5 91*33.17 165 28 28 28 70 1c 28 .16 420 77 77 49 28 2* 36 .15 515 54 47 54 10 25 3 5 .15 75 † † †

[0088] All patients had previously received one or more courses of all- trans retinoic acid, plus an plus arabinoside. *Denotes individuals with proven retinoid resistance (i.e. lack of response during reinduction a 15 or relapse while on retinoid maintenance); † Denotes patient who dies early, Other treatment:/ etoposide; ballogeneic bone marrow transplantation;cmethotrexate/vincristine/6-mercaptopurine; d9-cis retinoic acid plus M195 (anti-CD33 monoclonal antibody).

7. EXAMPLES: CLINICAL USE IN LYMPHOMA

20 [0089] Based upon the initial discovery of the antitumor effects of arsenic trioxidein vitro against B-cell lymphocyte lines, the inventors treated one patient with intermediate-grade large cell lymphoma who had relapsed from multiple forms of conventional therapy, including autologous bone marrow transplantation. Despite rapid progression of his disease prior to starting the arsenic trioxide therapy, treatment with arsenic trioxide effected a major (>50%) shrinkage

25 in the size of his cancerous lymph nodes and spleen, which was also associated with a major improvement of his quality of life.

8. EXAMPLES: CLINICAL USE IN NON-HEMATOPOIETIC CANCER

30 [0090] Arsenic trioxide was also used to treat colon cancer. In a preliminary test, one patient with colon cancer who received one treatment with arsenic trioxide showed a major reduction in his serum CEA (carcinoembryonic antigen) level. The patient received daily intravenous infusion of 0.1-5 mg arsenic trioxide per kg body weight per day for five days. A change in the level of CEA from 19,901 ng/ml to 15,266 ng/ml, a 23% reduction, was observed. It is well known that a reduced level of serum CEA is associated with antitumor response.

35 [0091] Clinical data confirms that arsenic trioxide can also be used to treat other non- hematopoietic cancer, such as colon cancer.

9. EXAMPLES: PHARMACOKINETICS STUDIES

40 [0092] Several dose-ranging studies were conducted to examine the pharmacokinetics (PK) and biological effects of As2O3 in patients with APL and in patients with other hematologic diseases. In patients with APL, marrow mononuclear cells were serially monitored by flow cytometry for immunophenotype, fluorescence in situ hybridization (FISH), and Western blot expression of the apoptosis-associated proteins, caspases 1, 2 and 3. Cells that coexpressed CD11b and CD33, and which by FISH analysis carried the (15;t 17) translocation, progressively increased during treatment and

45 persisted early in complete remission. As2O3 also induced in vivo expression of the proenzymes of caspase 2 and caspase 3, and activation of both caspase 1 and caspase 3. PK analysis of blood and urine for elemental arsenic (As) content showed that As was distributed in both plasma and fractions of whole blood. Parallel elimination curves suggested that these 2 compartments were freely exchangeable, and decayed from peak values with initial half lives of about 60 mins. The mean AUC on day 1 was about 400 ng hr/ml. Approximately 20% of the administered dose

50 was recovered in urine within the first 24 hrs. [0093] We then initiated a dose-ranging study in patients with diseases other than APL using a daily intravenous dosing schedule for a cumulative total of 25 days per treatment course every 3-5 weeks at dose levels of 0.1 and 0.15 mg per kg body weight per day. To date, 10 patients have been accrued, including patients with CLL (2 patients), AML (3 patients), lymphoma (4 patients), and CML (1 patient). Five patients were removed from the study early due to rapid

55 progression, and 5 completed the planned 25-day course. Over this dose range, the drug has proved well-tolerated; adverse effects have included skin rash, lightheadedness during the infusion, fatigue, and QTc prolongation on EKG.

Results from this ongoing study show that clinical use of As 2O3 induces partial differentiation and apoptosis in APL, but that the therapeutic effects of this agent are not confined to this disorder.

13 EP 1 964 557 B1

Claims

1. A method for preparing a pharmaceutical composition which comprises arsenic trioxide comprising the steps of:

5 a. solubilising arsenic trioxide in an aqueous solution at pH greater than 12 to form an arsenic trioxide solution; b. neutralizing the arsenic trioxide solution with an acid to a pH of 8 to 8.5; c. diluting the arsenic trioxide solution from step (b) in a 5% dextrose solution that stabilizes and lowers the pH to about 7.0; and d. sterilizing the pharmaceutical composition. 10

Patentansprüche

1. Verfahren zur Herstellung einer Arsentrioxid enthaltenden pharmazeutischen Zusammensetzung, welches die fol- 15 genden Schritte umfasst:

a. Lösen des Arsentrioxids in einer wässrigen Lösung bei einem pH-Wert von über 12 unter Bildung einer Arsentrioxidlösung; b. Neutralisieren der Arsentrioxidlösung mit einer Säure auf einen pH-Wert von 8 bis 8,5; 20 c. Verdünnen der Arsentrioxidlösung aus Schritt (b) mit einer 5%igen Dextroselösung, die stabilisiert und den pH-Wert auf etwa 7,0 senkt; und d. Sterilisieren der pharmazeutischen Zusammensetzung.

25 Revendications

1. Méthode de préparation d’une composition pharmaceutique qui comprend du trioxyde d’arsenic comprenant les étapes consistant à :

30 a. solubiliser du trioxyde d’arsenic dans une solution aqueuse à un pH supérieur à 12 pour former une solution de trioxyde d’arsenic ; b. neutraliser la solution de trioxyde d’arsenic avec un acide jusqu’à un pH de 8 à 8,5 ; c. diluer la solution de trioxyde d’arsenic de l’étape (b) dans une solution de dextrose à 5 % qui stabilise et réduit le pH à environ 7,0 ; et 35 d. stériliser la composition pharmaceutique.

40

45

50

55

14 EP 1 964 557 B1

REFERENCES CITED IN THE DESCRIPTION

This list of references cited by the applicant is for the reader’s convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description

• ROBBINS ; ANGELL. Basic Pathology. W.B. Saun- • GOODMAN ; GILMAN’S. The Pharmacological Ba- ders Co, 1976, 68-79 [0003] sis of Therapeutics. 1997, 1659-1662 [0009] • ROITT, I. ; BROSTOFF, J ; KALE, D. Immunology. • CHEN et al. Manual of Clinical Drugs. Shanghai In- 1993, 17.1-17.12 [0003] stitute of Science and Technology, 1995, 830 [0010] • CUTLER ; BRADFORD. Am. J. Med. Sci., 81-84 • MERVIS. Science, 1996, vol. 273, 578 [0010] [0008] • SHEN et al. Blood, 1997, vol. 89, 3354-3360 [0010] • FORKNER ; SCOTT. J. Am. Med. Assoc., 1931, vol. • CHEN et al. Blood, 1996, vol. 88, 1052-1061 [0010] iii, 97 [0008] • ANDRE et al. Exp. Cell Res., 1996, vol. 229, 253-260 • STEPHENS ; LAWRENCE. Ann. Intern. Med., 1936, [0010] vol. 9, 1488-1502 [0008] • ZHU et al. Proc. Natl. Acad. Sci., 1997, vol. 94, • MONFARDINI et al. Cancer, 1973, vol. 31, 492-501 3978-3983 [0010] [0008] • KIRK-OTHMER. Encyclopedia of Chemical Technol- • HUTCHINSON. Trans. Path. Soc. Lond., vol. 39, 352 ogy. John Wiley & Sons, vol. 3, 633-655 [0019] [0009] •ZHUal.et Proc. Natl. Acad. Sci. USA, 1997, vol. 94, • NEUBAUER. Br. J. Cancer, 1947, vol. 1, 192 [0009] 3978-3983 [0028] • CUZICK et al. Br. J. Cancer, 1982, vol. 45, 904-911 • KONIG et al. Blood, 1997, vol. 90, 562-570 [0028] [0009] • GOODMAN ; GILMAN’S. The Pharmacological Ba- • KASPAR et al. J. Am. Med. Assoc., 1984, vol. 252, sis of Therapeutics. 1997, 1660 [0033] 3407-3408 [0009] • WANG et al. Blood, 1998, vol. 92, 1497-1504 [0045] • LEE et al. Science, 1988, vol. 241, 79-81 [0009] • MILLER et al. Proc. Natl. Acad. Sci., 1992, vol. 89, •GERMOLECetal. Toxicol. Applied Pharmacol., 2694-8 [0073] 1996, vol. 141, 308-318 [0009] • MILLER et al. Blood, 1993, vol. 82, 1689-94 [0073]

15